Jan Schultke

I think the rules for this would be pretty complicated. For example, if you use `std::enable_if_t` in the return type, or some other traits or forms of doing SFINAE, trailing...

> #2 seems easily doable for the short list I found following your link... I swear we've talked about this before... maybe double check `cpp` to see if we haven't...

The whitelist would be quite short:  All other typedefs are covered by the rule for `_t` suffixed identifiers being highlighted as types.

I have added the missing include. This issue should be resolved now.

Consider the following issues with using by-value range-for variables: ```cpp // Scenario (1): the author does modify 'x' within the loop, so they intentionally make 'x' mutable. // This conforms...

The problem is that even though it is "more correct" to make such functions `[[nodiscard]]`, it's also very verbose. It's the same reason why recommending `const` on function parameters is...

> And the Right Thing is to use it widely, everywhere that it makes sense. The question is whether this makes sense as a guideline though. Afaik, the guidelines also...

It's difficult to encapsulate all of these scenarios in a single sentence, and that's ultimately what we need for a guideline. I would approach this in reverse; i.e. starting at...

> function call to get the length of the list - O(n) Are you confused by `sizeof...(Tail) == 0`? This is a constant expression, it doesn't have any runtime cost....

> > There is always a penalty. Its just a matter of where. > > So where is the penalty here? The penalty is that you will have to split...